Error pages

[doug]

Hi,
Does anyone know if theres a way that I can see what pages the visitors are going to, if they go to any of my old pages that used to be in circulation on the web please?

Doug.

The statcounter wasn't on any of the old pages - so I gues there's no way that I can see which of the old pages they are visiting?

[webado]

The statcounter wasn't on any of the old pages - so I gues there's no way that I can see which of the old pages they are visiting?
No. You can only check the web hosting stats if they are still around and accessible.

But I guess you're referring to people visiting your long-gone pages which may be cached. Then it's no.

[doug]

It's always good to ask :lol: You never know what type of magic is available :P

[robinev]

Maybe this isn't what you're looking for, but StatCounter will display requests for non-existent pages if you set up a custom error page that includes the StatCounter code.

I have several such entries in my current log. They appear because Google hasn't updated its image index in a long time and because of that, it includes dozens of outdated image url's for my site.

I can see in the StatCounter log exactly which of the image url's are requested from Google's outdated index because the non-existent page's url is displayed rather than the url of the error page (which is what folks actually see).

Yahoo has a similar problem, but their outdated url's are included in their main index -- an even more problematic bug.

I handle Yahoo's nasty buggyness with redirects whenever I see a bad request from their index (which is most of the time with them). On the image url's, I'd like to do more than just present the error page, but haven't yet been able to do it with my host's problematic support for php. But at least I can see in the StatCounter logs how often the problem arises.

Bottom line: If you include the StatCounter code on a custom error page, you'll be able to find out what outdated page a user is requesting. I don't think it's magic, but it seems to be an unintentional "feature".

[webado]

That's quite brilliant! It will work for all requests within a web site that still exists. The cached pages which may be accessed though won't trigger anything.

I think I might want to implement it myself on my custom error pages too. Thanks for the great tip. :D

Some time ago I noticed in Awstats some activity resulting in lots of 404 error codes (missing page requests). All the attempts were for various possible mail scripts, as if somebody was trying hard to find an unprotected mail script to use for spam or other such things. Luckily they didn't find any such scripts, but it got me thinking about securing everything as tightly as possible. Awstats shows the error, but not the IP of whomever tried that missing page, nor the date or time. I'd have liked to see that.

[doug]

Maybe this isn't what you're looking for, but StatCounter will display requests for non-existent pages if you set up a custom error page that includes the StatCounter code.

I have several such entries in my current log. They appear because Google hasn't updated its image index in a long time and because of that, it includes dozens of outdated image url's for my site.

I can see in the StatCounter log exactly which of the image url's are requested from Google's outdated index because the non-existent page's url is displayed rather than the url of the error page (which is what folks actually see).

Yahoo has a similar problem, but their outdated url's are included in their main index -- an even more problematic bug.

I handle Yahoo's nasty buggyness with redirects whenever I see a bad request from their index (which is most of the time with them). On the image url's, I'd like to do more than just present the error page, but haven't yet been able to do it with my host's problematic support for php. But at least I can see in the StatCounter logs how often the problem arises.

Bottom line: If you include the StatCounter code on a custom error page, you'll be able to find out what outdated page a user is requesting. I don't think it's magic, but it seems to be an unintentional "feature".

Many thanks for that, it could be just what I need.

One thing, (isn't there always 'just one thing' :roll: ) - on some of the old pages, there wasn't any images, would it still work?

[webado]

Many thanks for that, it could be just what I need.

One thing, (isn't there always 'just one thing' :roll: ) - on some of the old pages, there wasn't any images, would it still work?

The custom error pages are not about missing images. They are about currently missing pages altogether. So if your old pages still exist they will not register as missing. If they don't have the counter on them, you'll never know anybody accessed them. If the pages don't exist and their url is for the same domain where you have a custom error page for code 404 with counter code on it, then this is what will be logged. If the pages were on some other domain long gone and the visitor gets them out of the search engine's cache, then there's nothing you can do to trap that fact.

[robinev]

One thing, (isn't there always 'just one thing' :roll: ) - on some of the old pages, there wasn't any images, would it still work?

Yup. What Christina said.

I probably confused things by mentioning my image files (which are actually standard html pages that include an image). But as Christina pointed out, the 404 error page is most often called when a user requests an html file that doesn't exist on the server.

The log will show whatever the user entered as their request.